[wirelesstoronto-discuss] Mesh, mesh and more meshing about.

[Nicholas Kovats]

One single radio + antenna per Meraki if used as a single default gateway
(e.g. internet) will have a hard time "servicing" multiple  simultaneous
requests from multiple single radio Meraki nodes.

I second a test of the batman routing protocol but with multiple radio
boxes, e.g. openwrt/batman w/ 2-3 radios per embedded Linux box, i.e.

1. Each node would utilize two radios dedicated to lets say a backhaul 802.a
5.0 GHZ link (directional antennas)  +  one 802.n or 802.g radio servicing
local wireless clients with a mimo or omnidirectional antenna.

But the first backhaul radio could aim at the default gateway whilst the
second backhaul radio could aim at the next mesh node. They could be daisy
chained repeatedly without the crappy throughput dropping by 1/2  in the
more common WDS "meshes" that now exist.

2. Multiple Internet gateways could be utilized in this type of mesh and the
batman routing protocol will intelligently route a clients request to it's
nearest "default" internet gateway. With minimal overhead. The masses could
really  start "sharing" their multiple gateway connections.

3. You could potentially  "aggregate" the multiple internet gateway to
increase throughput in parts of your mesh.

4. Keep in mind spectrum, bandwidth, throughput are precious as they exist
in the reality of community mesh. You could intelligently and
"democratically" allocate bandwidth per application with simple traffic
engineering as opposed to the Evil Rogers Methodology of deep packet
inspection and throttling back your apps according to their corporate
agenda.

5. People tend to imagine mesh with movement of the wireless nodes. That is
very difficult to do in terms of seamless handoffs. The military is very
interested in a true mobile mesh but it would amazing for the WT wireless
mobile nodes knapsack idea.

6. A more practical approach are fixed wireless mesh nodes as per the above
description.

Let 'er rip..

NK

On Fri, Apr 18, 2008 at 1:09 PM, j.roks <aneasyone at gmail.com> wrote:

> testing range and repeat with meraki. I think gabe has some lying around.
> does anyone have an openmesh box? maybe WT can order a few open meshes and
> test the following Sunday instead.
>
>
> =j
>
> On 18-Apr-08, at 11:51 AM, "Nicholas Kovats" <nkovats at gmail.com> wrote:
>
> What do you have in mind, J?
>
> On Fri, Apr 18, 2008 at 10:58 AM, j.roks < <aneasyone at gmail.com>
> aneasyone at gmail.com> wrote:
>
> > anyone up for testing mesh distance ranges some time this weekend .
> > sunday?
> >
> > =j
> >
> > _______________________________________________
> > wirelesstoronto-discuss mailing list
> >  <wirelesstoronto-discuss at wirelesstoronto.ca>
> > wirelesstoronto-discuss at wirelesstoronto.ca
> > <http://lists.wirelesstoronto.ca/cgi-bin/mailman/listinfo/wirelesstoronto-discuss>
> > http://lists.wirelesstoronto.ca/cgi-bin/mailman/listinfo/wirelesstoronto-discuss
> >
>
> _______________________________________________
> wirelesstoronto-discuss mailing list
> wirelesstoronto-discuss at wirelesstoronto.ca
>
> http://lists.wirelesstoronto.ca/cgi-bin/mailman/listinfo/wirelesstoronto-discuss
>
>
> _______________________________________________
> wirelesstoronto-discuss mailing list
> wirelesstoronto-discuss at wirelesstoronto.ca
>
> http://lists.wirelesstoronto.ca/cgi-bin/mailman/listinfo/wirelesstoronto-discuss
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.wirelesstoronto.ca/pipermail/wirelesstoronto-discuss/attachments/20080418/5b4ffd3a/attachment-0001.htm